Deep learning-based prediction of one-year mortality in Finland is an accurate but unfair aging marker

IF 17 Q1 CELL BIOLOGY Nature aging Pub Date : 2024-06-24 DOI:10.1038/s43587-024-00657-5
Andrius Vabalas, Tuomo Hartonen, Pekka Vartiainen, Sakari Jukarainen, Essi Viippola, Rodosthenis S. Rodosthenous, Aoxing Liu, Sara Hägg, Markus Perola, Andrea Ganna
{"title":"Deep learning-based prediction of one-year mortality in Finland is an accurate but unfair aging marker","authors":"Andrius Vabalas, Tuomo Hartonen, Pekka Vartiainen, Sakari Jukarainen, Essi Viippola, Rodosthenis S. Rodosthenous, Aoxing Liu, Sara Hägg, Markus Perola, Andrea Ganna","doi":"10.1038/s43587-024-00657-5","DOIUrl":null,"url":null,"abstract":"Short-term mortality risk, which is indicative of individual frailty, serves as a marker for aging. Previous age clocks focused on predicting either chronological age or longer-term mortality. Aging clocks predicting short-term mortality are lacking and their algorithmic fairness remains unexamined. We developed a deep learning model to predict 1-year mortality using nationwide longitudinal data from the Finnish population (FinRegistry; n = 5.4 million), incorporating more than 8,000 features spanning up to 50 years. We achieved an area under the curve (AUC) of 0.944, outperforming a baseline model that included only age and sex (AUC = 0.897). The model generalized well to different causes of death (AUC > 0.800 for 45 of 50 causes), including coronavirus disease 2019, which was absent in the training data. Performance varied among demographics, with young females exhibiting the best and older males the worst results. Extensive prediction fairness analyses highlighted disparities among disadvantaged groups, posing challenges to equitable integration into public health interventions. Our model accurately identified short-term mortality risk, potentially serving as a population-wide aging marker. Here the authors show that deep learning accurately predicts one-year mortality using nationwide Finnish data. Despite robust performance and potential as an aging marker, fairness analyses reveal prediction disparities, urging cautious integration into public health.","PeriodicalId":94150,"journal":{"name":"Nature aging","volume":"4 7","pages":"1014-1027"},"PeriodicalIF":17.0000,"publicationDate":"2024-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.nature.com/articles/s43587-024-00657-5.pdf","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Nature aging","FirstCategoryId":"1085","ListUrlMain":"https://www.nature.com/articles/s43587-024-00657-5","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"CELL BIOLOGY","Score":null,"Total":0}
引用次数: 0

Abstract

Short-term mortality risk, which is indicative of individual frailty, serves as a marker for aging. Previous age clocks focused on predicting either chronological age or longer-term mortality. Aging clocks predicting short-term mortality are lacking and their algorithmic fairness remains unexamined. We developed a deep learning model to predict 1-year mortality using nationwide longitudinal data from the Finnish population (FinRegistry; n = 5.4 million), incorporating more than 8,000 features spanning up to 50 years. We achieved an area under the curve (AUC) of 0.944, outperforming a baseline model that included only age and sex (AUC = 0.897). The model generalized well to different causes of death (AUC > 0.800 for 45 of 50 causes), including coronavirus disease 2019, which was absent in the training data. Performance varied among demographics, with young females exhibiting the best and older males the worst results. Extensive prediction fairness analyses highlighted disparities among disadvantaged groups, posing challenges to equitable integration into public health interventions. Our model accurately identified short-term mortality risk, potentially serving as a population-wide aging marker. Here the authors show that deep learning accurately predicts one-year mortality using nationwide Finnish data. Despite robust performance and potential as an aging marker, fairness analyses reveal prediction disparities, urging cautious integration into public health.

Abstract Image

Abstract Image

查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
基于深度学习的芬兰一年死亡率预测是一个准确但不公平的老龄化标志。
短期死亡风险表明个人体质虚弱,是衰老的标志。以前的年龄钟主要预测计时年龄或长期死亡率。目前还缺乏预测短期死亡率的年龄钟,其算法的公平性也尚未得到研究。我们开发了一个深度学习模型,利用芬兰全国人口纵向数据(FinRegistry;n = 540 万)预测 1 年死亡率,其中包含 8000 多个特征,时间跨度长达 50 年。我们的曲线下面积(AUC)达到了 0.944,优于仅包含年龄和性别的基线模型(AUC = 0.897)。该模型能很好地概括不同的死因(50 种死因中有 45 种的 AUC > 0.800),包括训练数据中没有的 2019 年冠状病毒疾病。不同人群的预测结果各不相同,年轻女性的预测结果最好,而老年男性的预测结果最差。广泛的预测公平性分析凸显了弱势群体之间的差异,这给公平纳入公共卫生干预措施带来了挑战。我们的模型能准确识别短期死亡风险,有可能成为全人口老龄化的标志。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 去求助
来源期刊
CiteScore
14.70
自引率
0.00%
发文量
0
期刊最新文献
Boosting NAD+ in patients with COPD reduces airway inflammation. Dysregulation of endogenous retroviruses triggers aging and senescence. Effect of nicotinamide riboside on airway inflammation in COPD: a randomized, placebo-controlled trial. Reactivation of senescence-associated endogenous retroviruses by ATF3 drives interferon signaling in aging. Brain-wide alterations revealed by spatial transcriptomics and proteomics in COVID-19 infection
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1